|
This statement is now a couple of years out of date. To read something
more recent, please look at my July 2008 report to the NSF.
I have a longstanding interest in stellar endpoints, particularly in
observational constraints on the neutron star equation of state, and
in supernova remnants. I started my research career working on
dual-frequency observations of pulsars and on the radio expansion of
Cassiopeia A. I am now leading searches for isolated neutron
stars and for millisecond pulsar companions to low-mass white dwarfs,
and I am also involved in X-ray studies of the supernova remnant G16.7+0.1.
Much of my current work exploits the information gleaned from
correlations of data from the optical Sloan Digital Sky Survey with
that from surveys at other wavelengths.
- Contents:
- Optical Identifications of X-ray Sources
- "Ordinary" Stellar Counterparts
- Unusual Stellar Counterparts
- Extraordinary Stellar Counterparts: Isolated Neutron Stars
RASS/SDSS Active Galactic Nuclei
- Millisecond Pulsar Companions to Low-Mass White Dwarfs
- Other Multiwavelength Work
- Supernova Remnants
- Correlations of the SDSS and GALEX Catalogs
- References
Optical Identifications of X-ray Sources
My Ph.D. thesis work is part of an effort to produce a complete
sample of X-ray sources included in the ROSAT All-Sky Survey (RASS)
from correlations with the Sloan Digital Sky Survey (SDSS).
X-ray data alone often cannot determine whether a source is
Galactic or extragalactic, much less make finer distinctions about
its nature. Yet the optical counterparts to the > 100,000 RASS
sources form a heterogeneous group of objects ranging from distant
quasars to nearby active M dwarfs (e.g., Zickgraf et al. 2003).
The identification of optical counterparts is therefore an essential
companion study to large X-ray surveys.
"Ordinary" Stellar Counterparts
Galactic stars account for about one-third of all RASS
identifications. Aside from cataclysmic variables (CVs) and equally rare
white dwarfs (WDs), however, the vast majority of stellar X-ray emitters
cataloged in the RASS are unlikely to be discovered from routine SDSS
spectroscopy. Normal stars that are luminous in the X ray relative
to their optical output are almost all optically it brighter than
the m = 15 SDSS spectroscopic limit (e.g., Stocke et al. 1991).
For my Ph.D. thesis, I have therefore
used the Dual Imaging Spectrograph on the 3.5-m telescope at Apache Point
Observatory, NM, to obtain spectra for more than 700 bright stars
detected in SDSS and the Two Micron All Sky Survey (2MASS) near a
RASS source. An SDSS or 2MASS magnitude is used to estimate
f_x/f_opt ratios for these stars; the ratios are
then compared to those of known X-ray emitters of the same spectral type,
thereby verifying that these stars are indeed plausible RASS source
counterparts. Most of the identified counterparts
are late-type stars with coronal emission, but I have also discovered
a number of other interesting objects, including candidate WDs and
CVs.
Unusual Stellar Counterparts
I have contributed to a series of papers (Szkody et
al. 2002, 2004, 2005) that include descriptions of the X-ray properties of
SDSS CVs. I have also collected the SDSS spectra for
a few dozen SDSS X-ray-emitting WDs. In collaboration with
Bruce Margon and Ron Downes (STScI), I am currently conducting a search for
unexpected spectral features in these WDs by stacking their
spectra and comparing the obtained composite spectra to those from the much
larger sample of "ordinary" WDs.
Extraordinary Stellar Counterparts: Isolated Neutron Stars
Stellar evolution predicts that the number of detected pulsars
in the Galaxy (2000) should be dwarfed by the number of older,
inactive neutron stars (NSs), most of which are too cool to be visible in the
X ray and rotate too slowly to generate radio pulses
(e.g., Kulkarni & van Kerkwijk 1998). However, some of these stars may
reheat by accreting material from the interstellar
medium, and ROSAT was expected to detect hundreds to thousands of
rejuvenated NSs (e.g., Treves & Colpi 1991). These isolated neutron
stars (INSs) were predicted to have blackbody spectra that would
place strong constraints on the NS equation of state. Yet only seven INSs
are currently known (the "Magnificent Seven"), a sample that manages to
be both too small and too diverse in detail to address the NS equation of
state definitively. Clearly, obtaining a larger sample is highly
desirable, and as part of my thesis work I have used SDSS to search for
new INS candidates in the RASS.
Four of the Magnificent Seven have suggested optical counterparts with
m_V ≥ 25 (Kaplan, Kulkarni, & van Kerkwijk 2003), and new INS
counterparts are therefore likely to be too faint to be found directly
using SDSS. Instead, I developed an algorithm to search for RASS sources
devoid of plausible optical counterparts to the SDSS
≈ 22 mag limit. These are sources with such extreme f_x / f_opt
ratios that an INS becomes a plausible identification.
This search is an order of magnitude more selective than previous ones;
it excludes 99.9% of the initial RASS sources considered
(Agüeros et al. 2006).
Nine new candidates were found, and I have an upcoming
Chandra X-ray Observatory program to confirm the nature of these sources
(see Fig. 1 for SDSS images of three of my INS candidates). I will
lead the follow-up observations required to
explore the physical properties of these objects (because the candidate
INSs all have extreme f_x / f_opt ratios, even those that are
it not INSs will be highly interesting and warrant further study).
 |
| Fig. 1: SDSS images of three INS candidates to be observed by Chandra, plus the field of the known INS RX J1605.3+3249 (far right), also recovered by my algorithm. The images are 1' across and are shown with the same optical stretch to allow comparison. Each circle encloses an area bereft of plausible counterparts to the X-ray source. The circles are of radii 56'', 60'', 24'', and 28'', corresponding to four times the relevant RASS positional uncertainty; the it brightest optical object within any of the circles is g = 21.7 magnitudes. |
RASS/SDSS Active Galactic Nuclei
As part of the broader work of identifying it all the RASS
sources in SDSS, I also worked on
the identification of SDSS counterparts to X-ray-emitting quasars and
related active galactic nuclei (AGN). With about one-fifth of
the SDSS survey completed, we characterized more than 1200
ROSAT-detected AGN (Anderson et al. 2003).
Millisecond Pulsar Companions to Low-Mass White Dwarfs
Observations of field radio pulsars indicate that 5% reside
in binary systems, and that in most cases the companion is a low-mass
WD (0.1 to 0.4 M_solar). These pulsars are most often old,
spun up by accretion and "reborn" as millisecond pulsars (MSPs).
Such MSP/low-mass WD systems are valuable for exploring binary
evolution and the properties of their components, and have notably been
used to constrain the maximum NS mass, and hence the NS equation of state
(e.g., Bell et al. 1995). However, only half a dozen of these binary
systems have been exploited fully (see van Kerkwijk et al. 2005).
MSPs are found in radio surveys, and the companion
low-mass WDs are generally too faint for high-resolution
optical spectroscopy.
In collaboration with Fernando Camilo (Columbia), Scot
Kleinman (Subaru), and others,
I recently started a program to search for radio pulsar companions to
optically selected SDSS low-mass WDs. Less than 0.5% of the
6000 new spectroscopically confirmed SDSS WDs
have masses ≤ 0.3 M_solar
(Eisenstein et al. 2006). I am the PI on a recently
approved proposal to search for putative companions to two low-mass
WDs with the Green Bank Telescope (GBT). I have also submitted a
proposal to observe another 13 recently discovered SDSS low-mass
WDs--including two with possibly the lowest masses yet detected--with
the GBT. Because the selection is entirely different from the usual methods
for finding MSPs, we may obtain independent estimates
of their number and birthrate. In addition, these low-mass WDs
are all optically brighter than the known MSP companions, and high-resolution
spectroscopy will therefore be able to constrain the orbital
parameters of these systems to unprecedented precision.
Other Multiwavelength Work
Supernova Remnants
I have worked on several projects involving
radio and X-ray observations of Galactic supernova remnants (SNRs).
For my M.Phil. thesis, I used the Cambridge Low-Frequency Synthesis
Telescope to measure the bulk expansion at 151 MHz of Cassiopeia A,
the youngest known Galactic SNR. I found that the derived dynamical age was
consistent with Cas A being in the transition between the
free expansion and Sedov-Taylor (adiabatic) evolutionary stages, contradicting
previous radio studies. However, those studies relied on measurements
of individual radio features, while the age I obtained was consistent
with that derived from X-ray studies that were also sensitive
to Cas A's bulk expansion
(Agüeros & Green 1999).
More recently, I worked on XMM-Newton and Chandra observations
of G16.7+0.1, a classic composite SNR discovered with the Very
Large Array (Helfand et al. 1989) with one of the faintest radio core
components detected to date. XMM data were used to characterize the
remnant's pulsar wind nebula (Helfand, Agüeros, & Gotthelf 2003).
The Chandra data are currently being analyzed to reveal
the strength of the central pulsar and to explore the possibility that
the wind nebula has recently been crushed by the SNR's
reverse shock (Helfand et al., in preparation).
Correlations of the SDSS and GALEX Catalogs
Working with Željko Ivezić (Washington) and others,
I matched SDSS with the first publicly released data from the
ultraviolet (UV) sky survey conducted by the Galaxy Evolution Explorer
(GALEX). With the addition of 2MASS data, I
constructed 10-band UV-optical-infrared spectral energy
distributions for stars, WDs, low-redshift quasars, and galaxies.
While only 2.5% of SDSS sources are detected by
GALEX, I found that the UV data carry important information:
UV colors allow much more accurate and robust estimates of
star-formation rates than possible using only SDSS data, for example.
However, the UV spectral slope for most galaxies
detected in both GALEX bands is consistent both with hot stars and
with the presence of an AGN. Although previous studies interpreted the
GALEX detections of a sample of SDSS elliptical galaxies as evidence
for recent star formation, our work suggests that their UV
emission may instead reveal low-level AGN activity in these galaxies
(Agüeros et al. 2005; see Fig. 2).
 |
Fig. 2: Top: Normalized median ultraviolet-through-infrared spectral energy distributions (SEDs) for low-redshift quasars (z ~ 1), hot WDs, and turn-off stars constructed using GALEX, SDSS, and 2MASS data. The data points are connected to guide the eye. Bottom: Mean SEDs for low-redshift blue (spiral) and red (elliptical) galaxies. The observed UV slope for blue galaxies is consistent with the UV slope for both hot stars and low-redshift quasars (Agüeros et al. 2005).
|
References
Agüeros, M. A., & Green, D. A. 1999, MNRAS, 305, 957
Agüeros, M. A., et al. 2006, AJ, 131, 1740
--. 2005, AJ, 130, 1022
Anderson, S. F., et al. 2003, AJ, 126, 2209
Bell, J. F., et al. 1995, ApJ, 452, L121
Eisenstein, D. J., et al. 2006, ApJs, 167, 40
Helfand, D. J., et al. 1989, ApJ, 341, 151
Helfand, D. J., Agüeros, M. A., & Gotthelf, E. V. 2003, ApJ, 592, 941
Kaplan, D. L., Kulkarni, S. R., & van Kerkwijk, M. H. 2003, ApJ, 588, L33
Kulkarni, S. R., & van Kerkwijk, M. H. 1998, ApJ, 507, L49
Stocke, J. T., et al. 1991, ApJS, 76, 813
Szkody, P., et al. 2002, AJ, 123, 430
--. 2004, AJ, 128, 1882
--. 2005, AJ, 129, 2386
Treves, A., & Colpi, M. 1991, A&A, 241, 107
van Kerkwijk, M. H., et al. 2005, in ASP Conference Series Vol. 328, ed. F. A. Rasio & I. H. Stairs, 357
Zickgraf, F.-J., et al. 2003, A&A, 406, 535
|
